We use a Monte Carlo approach to explore the potential impact of observation and inversion model errors on the spatial statistics of field-estimated unsaturated hydraulic properties. For this analysis we simulate tension infiltrometer measurements in a series of idealized realities, each consisting of spatially correlated random property fields. We consider only simple measurement errors that can be easily modeled. We show that estimated hydraulic properties are strongly biased by small, simple observation and inversion model errors. This bias can lead to order-of-magnitude errors in spatial statistics and artificial cross correlation between measured properties. The magnitude of bias varies with the true mean of the property field, the type of error considered, and the type of spatial statistic. We find no unique indicators of bias as property values may appear reasonable and spatial statistics may look realistic. Our results suggest new concerns for geostatisticians, stochastic modelers, and unsaturated zone practitioners who are unaware of the potential impact of spatial bias in field-estimated properties.